When checking a plurality of documents, it is desirable to know which of the documents is important before the check. For example, a business manager who should routinely check many contracts wants to check contracts in the order of importance. Therefore, techniques are being considered to evaluate how important documents such as contracts are.
Patent Document 1 describes a document search method that uses an artificial intelligence model to score documents and select a subset of the highest scored documents.
Patent Document 2 describes a contract analysis system that generates a document vector for each clause on input contract data, compares the document vector for each clause with a document vector for each clause group to specify a legal provision included in the similar clause group as a related provision of the contract clause, and generates an analysis result screen that lists the related provisions for each contract clause.